PDA

View Full Version : Emergency help..


kelvintc
Aug12-04, 08:40 AM
I encountered many problems while doing exercises in text books. :confused: And i have stated it down in a word document attached in this post. Hope someone can help and teach me how to solve those problems. Answers are given. I just don't know how to get those answer.

Thanks a lot. :smile:

p/s : Emergency = next week test. :bugeye:

(2), (4), (5), (6) solved.. thanks

arildno
Aug12-04, 08:51 AM
Welcome to PF!

You must show some of your own work here; don't expect your homework to be done for you!
Having said that, let's take a specific question, nr. 4:
Now, what are your problems with this particular exercise?
Make a detailed comment on this.

kelvintc
Aug12-04, 09:03 AM
r dot grad T should be a scalar ? i wonder y answer's a vector. I used formulas and can't get those answers. Just can't understand why. So seeking help. Thanks.

arildno
Aug12-04, 09:11 AM
The answer should be a vector, as it is given in the answer (i haven't checked if the answer given is correct)

I hope you know about the grad "vector":
grad=\vec{a}_{x}\frac{\partial}{\partial{x}}+\vec{ a}_{y}\frac{\partial}{\partial{y}}+\vec{a}_{z}\fra c{\partial}{\partial{z}}

For example, the divergence of a vector \vec{v} is given by:
div\vec{v}=grad\cdot\vec{v}

Are you familiar with this notation?

kelvintc
Aug12-04, 09:20 AM
actually i dunno what means (r dot grad) ? i know div and grad as well
p/s: how you draw those symbols?

arildno
Aug12-04, 09:36 AM
You can click on the LATEX code to see how you write it .

OK, so you know the "grad", which I'll henceforth write as \nabla

Let's first review how we get the scalar known as "divergence"
(\nabla\cdot\vec{v})
Let \vec{v}=u\vec{a}_{x}+v\vec{a}_{y}+w\vec{a}_{z}

We then have that:
\nabla\cdot\vec{v}=\vec{a}_{x}\cdot(\frac{\partial }{\partial{x}}\vec{v})+\vec{a}_{y}\cdot(\frac{\par tial}{\partial{y}}\vec{v})+\vec{a}_{z}\cdot(\frac{ \partial}{\partial{z}}\vec{v})

This simplifies to, in our case:
\nabla\cdot\vec{v}=\frac{\partial{u}}{\partial{x}} +\frac{\partial{v}}{\partial{y}}+\frac{\partial{w} }{\partial{z}}

(Please comment if this doesn't make sense to you!)

Now, we're ready to tackle \vec{v}\cdot\nabla
This is also a "dot" product (scalar product), and looks like:
\vec{v}\cdot\nabla=u\vec{a}_{x}\cdot\nabla+v\vec{a }_{y}\cdot\nabla+w\vec{a}_{z}\cdot\nabla

Or, simplified:
\vec{v}\cdot\nabla=u\frac{\partial}{\partial{x}}+v \frac{\partial}{\partial{y}}+w\frac{\partial}{\par tial{z}}

This is a "scalar" operator which you then apply on T.

arildno
Aug12-04, 09:55 AM
The answer is correct, BTW

kelvintc
Aug12-04, 09:57 AM
oic thanks

kelvintc
Aug12-04, 10:04 AM
in q 5 : i found the surface integral = -2, but line integral = 7/6. it didn't match Stoke's theorem.. i wonder y....

kelvintc
Aug12-04, 10:05 AM
i never learn that inverse dot product b4.. hehe... thanks

kelvintc
Aug12-04, 10:07 AM
and q 2 i really have no idea

arildno
Aug12-04, 10:18 AM
OK, first:
Have you checked that you get no.4 right?

Secondly, try and group together a few questions you think "belong" to each other, which you would like to focus on.

kelvintc
Aug12-04, 10:25 AM
no.1 to 6 are about vectors. All about integrals.. I just wonder y can't get answers using formulas. No.2 is really dunno how to start also.

kelvintc
Aug12-04, 10:27 AM
no.5 curl = -x^2 in z direction.. then integral can't get 7/6

arildno
Aug12-04, 10:34 AM
OK, we'll look into 2 (but you did check 4, or what?).

2.
Now, you've been given equations for two surfaces.
In general, if you have a surface given on the form S(x,y,z)=0 (or constant),
you know that the normal on that surface at a point (x,y,z) is parallell to the gradient of S (evaluated on the same point).
Post what you get here in some detail.

kelvintc
Aug12-04, 10:35 AM
can't get no.4 answer

arildno
Aug12-04, 10:36 AM
Post what you've done. In detail.

kelvintc
Aug12-04, 11:13 AM
1) since z=0, a_{r} has no a_{x} component , a_{\phi} = -\sin{\phi}{a}_{r} + \cos{\phi}{a}_{\phi} ... and i can't t the answer for (a) and so on

arildno
Aug12-04, 11:17 AM
I meant on Q4; the one I started with.

kelvintc
Aug12-04, 11:26 AM
(r dot grad T ) = (\vec{r}\cdot\nabla){T}={{x}a_{x}}\frac{\partial{2 zy}}{\partial{x}}+{{y}a_{y}}\frac{\partial{xy^2}}{ \partial{y}}+{{z}a_{z}}\frac{\partial{x^2yz}}{\par tial{z}}
is it?

kelvintc
Aug12-04, 11:33 AM
(2), (3) no ideas
(5) i got \nabla\times{F}={-x^2}{a}_{z} and can't get the answer

arildno
Aug12-04, 11:33 AM
Absolutely not!
We gained:
\vec{r}\cdot\nabla=x\frac{\partial}{\partial{x}}+y \frac{\partial}{\partial{y}}+z\frac{\partial}{\par tial{z}}

We then have:
(\vec{r}\cdot\nabla)\vec{T}=x\frac{\partial\vec{T} }{\partial{x}}+y\frac{\partial\vec{T}}{\partial{y} }+z\frac{\partial\vec{T}}{\partial{z}}

arildno
Aug12-04, 11:36 AM
2) Take your first surface (given as an equation)
a)Rewrite that equation into a form S(x,y,z)=0,(introduce S(x,y,z) for the expression in x,y,z)

b) Calculate the gradient of S

kelvintc
Aug12-04, 11:39 AM
ooops.. finally understand and get it.. thanks for patience... hehe

kelvintc
Aug12-04, 11:41 AM
2) Take your first surface (given as an equation)
a)Rewrite that equation into a form S(x,y,z)=0,(introduce S(x,y,z) for the expression in x,y,z)

can't get it.... for an example?

arildno
Aug12-04, 11:41 AM
In question 2, the coordinates for the intersection doesn't make sense

Either it should be (1,2,1) (not (1-2,1)), or there is some other wrong troubling it.

Question 3 doesn't seem to make any sense at all..

kelvintc
Aug12-04, 11:44 AM
oh typed it wrong.. should be (-1, 2, 1)

(3) it wants \oint{V}{dS}

kelvintc
Aug12-04, 11:48 AM
that means S1(x,y,z)={x^2}{y}+{z-3}, S2(x,y,z)={x}{\log}{z}-{y^2}+{4} ?
but {\nabla}{S2} i never learn b4 (the log one)

arildno
Aug12-04, 11:52 AM
Allright, I'll give you an example of what I mean:
(x-a)^{2}+(y-b)^{2}+(z-c)^{2}=R^{2}

Clearly, this equation describes the shell of a sphere with center at (a,b,c).
We may rewrite the equation as:
S(x,y,z)=0
where in this case, we have:
S(x,y,z)=(x-a)^{2}+(y-b)^{2}+(z-c)^{2}-R^{2}

That is, the spherical shell is composed of those points on which the function S is zero(right?)

The gradient of S is easily found:
\nabla{S}=2(x-a)\vec{a}_{x}+2(y-b)\vec{a}_{y}+2(z-c)\vec{a}_{z}

The unit normal at a given point (x,y,z) is parallell to \nabla{S} there, but of unit length.
See if this helps you along.

arildno
Aug12-04, 11:54 AM
You have the correct expressions for S1 and S2.
Now differientiate with the gradient; log(z) is the natural logarithm to z, if you are familiar with that concept.

kelvintc
Aug12-04, 12:00 PM
{S1}{(x,y,z)}={x^2}{y}+{z-3}
{\nabla}{S1}={2xy}a_{x}+{x^2}a_{y}+a_{z}
S2(x,y,z)={x}{\log}{z}-{y^2}+{4}
{\nabla}{S2}={\log}{z}a_{x}-{2y}a_{y}+a_{z}\frac{{x}{\log}{e}}{z} ?
{\cos}{\phi}=\frac{{\nabla}{S1}{\cdot}{\nabla}{S2} }{{\mid}{\nabla}{S1}{\cdot}{\nabla}{S2}{\mid}} ?

arildno
Aug12-04, 12:08 PM
Precisely!
Now, you must figure out the unit normals at the intersection point, and calculate the angle between them. (log(e)=1, BTW)

kelvintc
Aug12-04, 12:16 PM
log e is not in base 10? it's in base e ? i get the answer correct if log e = 1.. :smile:

arildno
Aug12-04, 12:20 PM
Sure:
Now, the natural logarithm of a number "a" is a number b=log(a), so that:
e^{b}=a (right?)
If a=e, we have:
e^{b}=e=e^{1}\to{b}=log(e)=1

kelvintc
Aug12-04, 12:22 PM
ic.. thanks...

kelvintc
Aug12-04, 12:29 PM
oic.... about no.5 it asks me to prove the Stoke's Theorem.
i get \nabla\times{F}={-x^2}{a}_{z}
to find \int(\nabla\times{F})\cdot{dS} , do i need to seperate the triangle to 2 pieces, one is x=0 to x=1, y=0 to y=1 and another one is x=1 to x=2, y=1 to y=0 ?
I can't get 7/6 with both ways i did.
But, \oint{F}\cdot{dl} i got 7/6.. curious..

kelvintc
Aug12-04, 12:34 PM
i'm taking electrical engineering...

arildno
Aug12-04, 12:57 PM
1.We agree on the expression on the curl
2. I don't see the need to split up the triangle (I'll get into that)
3. I took a master's in fluid mechanics some time ago..

arildno
Aug12-04, 01:14 PM
My choices for the two triangles:
0<=x<=1,0<=y<=x triangle 1

1<=x<=2, 0<=y<=2-x triangle 2

kelvintc
Aug12-04, 10:30 PM
yet, i can't get the 7/6. my answer for triangle 1 = -1/4, then triangle 2 = -19/40... sums up can't get 7/6.. anything wrong?

master! oh, it's incredible..

kelvintc
Aug12-04, 10:42 PM
about (1) since z=0, a_{r} has no a_{x} component , a_{\phi} = -\sin{\phi}{a}_{x} + \cos{\phi}{a}_{y}
e^{-2}\sin\frac{\phi}{2}*-\sin\phi=-0.102

kelvintc
Aug12-04, 10:57 PM
no.6 i translate all x,y,z into cylindrical coordinates.
x = r\cos\phi
y = r\sin\phi
z = z
a_{x} = \cos{\phi}{a}_{r} - \sin{\phi}{a}_{\phi}
a_{y} = \sin{\phi}{a}_{r} + \cos{\phi}{a}_{\phi}
a_{x} = a_{z}
dS = rd\phi{d}{r}a_{r}
Hence,
F_{r} = r^2\cos^3\phi + r^2\sin^3\phi
integrate i can't get the answer also..

arildno
Aug13-04, 08:10 AM
Hi again!
I got 7/6 in 5); I'll take that first.

1. 7/6 is the number you get by the line integral going around the surface in one direction; -7/6 if you go in the opposite direction

2. We let the normal vector in the surface integral be \vec{n}=-\vec{a}_{z}

3. Hence, we must calculate I:
I=\int_{0}^{1}\int_{0}^{x}x^{2}dydx+\int_{1}^{2}\i nt_{0}^{2-x}x^{2}dydx

The first integral is: I_{1}=\frac{1}{4}
The second integral is: I_{2}=\frac{11}{12}

This yields 7/6 in total..

kelvintc
Aug13-04, 09:01 AM
thanks again.. hehe... oh... i really made too much mistakes in calculation.. sad

arildno
Aug13-04, 09:17 AM
No wonder you're stuck by 6)!

As far as I can see, the surface integral (the flux) on the cylindrical shell is 0.

However, if you consider the fluxes through the top and bottom disks (z=0 and z=2) as well, you'll get for the total flux 16\pi which seems to agree with the answer..

arildno
Aug13-04, 09:20 AM
Please post some work on 1)

kelvintc
Aug13-04, 09:53 AM
replace z=0 and z=2 to get (z^2-1) ?
then find \int{rd\phi}{dr} ?
by this i get 16\pi...
i also get the surface integral in r direction = 0

kelvintc
Aug13-04, 09:58 AM
(1) a_{r} = \cos{\phi}{a}_{x} + \sin{\phi}{a}_{y}
a_{\phi} = -\sin{\phi}{a}_{x} + \cos{\phi}{a}_{y}
a_{z} = a_{z}
then z=0, so vector in x direction only depends on phi .
-e^{-2}\sin\frac{\phi}{2}\sin\phi=-0.0586

arildno
Aug13-04, 10:09 AM
I'm not sure if your post 47 is a cry for help, or if you got my meaning.
I'll look into 1) later

kelvintc
Aug13-04, 10:14 AM
i think i got the answer for (6). just not sure whether the method is wrong or not

arildno
Aug13-04, 10:38 AM
Post in detail how you get the answer in 6)

kelvintc
Aug13-04, 10:53 AM
i got 4 for z=0 to z=2 for (z^2-1) then integrate get 2pie x r since r=2 then 4 times 4pie get 16pie

arildno
Aug13-04, 11:17 AM
Ok, you've got two disks you've got to integrate over in 6)

a) The disk placed at z=2.
At that plane, we have :
\vec{F}\cdot\vec{n}=(z^{2}-1)|_{(z=2)}=3
Or, it's contribution is: 3*\pi(2)^{2}=12\pi

b) The disk placed at z=0
At that plane, we have:
\vec{F}\cdot\vec{n}=(1-z^{2})|_{(z=0)}=1
Or, it's contribution is: \pi(2)^{2}=4\pi

Is this how you thought?

kelvintc
Aug14-04, 12:04 AM
ya... i did it like that

kelvintc
Sep3-04, 09:37 AM
i had been doing these qs over the past few weeks and there are still some questions remained unsolved.. anyone can help ?
q3, 7, 8, 9, 13, 14, 15